MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog Ser

Vol. 249: 223–236, 2003 Published March 10

INTRODUCTION

Two different water types control the hydrography ofthe surface ocean off the Oregon coast during summermonths. Continental shelf waters are dominated bysubarctic coastal water transported south from south-ern British Columbia and the Gulf of Alaska. Oceanicwaters offshore of the continental shelf represent aninshore arm of the California Current and originate tothe west, as part of the West Wind Drift and theso-called Transition Zone (Hickey 1998). The summerhydrography of shelf and slope waters is also influ-enced by wind-driven coastal upwelling, whereby

cold, salty, nutrient-rich, deep water is brought to thesurface where it displaces coastal and offshore waters,creating strong temperature and salinity gradients inboth continental shelf and slope waters (Hickey 1998).As the coastal waters move south along the Oregonand California coasts, they warm due to solar heatingand salinity rises as greater volumes of upwelled watermix with Transition Zone water (Hickey 1998).

Given the different origins of waters found off thecoastal northeast Pacific, we expected that each watertype would carry a different or unique zooplanktonassemblage into the region (Johnson & Brinton 1963).Moreover, given the onshore-offshore gradients in

© Inter-Research 2003 · www.int-res.com*Email: cheryl.morgan@noaa.gov

Onshore-offshore variations in copepodcommunity structure off the Oregon coast during

the summer upwelling season

Cheryl A. Morgan1,*, William T. Peterson2, Robert L. Emmett2

1Cooperative Institute for Marine Resources Studies, Oregon State University, Hatfield Marine Science Center,2030 S. Marine Science Dr., Newport, Oregon 97365, USA

2National Marine Fisheries Service, Northwest Fisheries Science Center, Hatfield Marine Science Center,2030 S. Marine Science Dr., Newport, Oregon 97365, USA

ABSTRACT: During the summer of 1994, 1996, and 1997, we conducted hydrographic and planktonsurveys of the upper 70 m in the nearshore, the continental shelf, and off-shelf oceanic waters offOregon, USA. Copepod densities and biomass were estimated along 4 transects from each cruise.The on-shelf copepod biomass was 2.8 times greater than off-shelf biomass. Nonmetric multidimen-sional scaling identified an on-shelf and an off-shelf copepod community. The change in communitycomposition usually occurred at or slightly offshore of the continental shelf break, defined here as180 m water depth. Indicator-species analysis identified the subarctic neritic species, Calanus mar-shallae, Pseudocalanus mimus, and Acartia longiremis, as good indicators of continental shelf waters.‘Warm water’ species Mesocalanus tenuicornis, Calocalanus styliremis, Clausocalanus spp., andCtenocalanus vanus were indicators of off-shelf waters. The copepod communities off the coast ofOregon during the summer upwelling season reflect the origins of the dominating currents withineach habitat. The coastal copepod community is subarctic neritic in origin, consistent with southwardcoastal flows. The offshore copepod community is a mixture of species with origins in the TransitionZone as well as species that are typical of the coastal region of the California Current off central andsouthern California. We speculate that the central and southern California Current species are pre-sent in offshore waters of Oregon in summer because they are transported north with the DavidsonCurrent in winter but, with the initiation of the upwelling season in spring, they are transported intooffshore waters where they establish viable populations.

KEY WORDS: Community structure · Copepod distributions · Nonmetric multidimensional scaling ·Oregon

Resale or republication not permitted without written consent of the publisher

Mar Ecol Prog Ser 249: 223–236, 2003

hydrographic conditions, we also expected strongcross-shelf zonation in zooplankton community struc-ture. Subarctic neritic species should dominate on-shelf waters, while oceanic, Transition Zone speciesshould dominate off-shelf waters.

Cross (1964) and Cross & Small (1967) were the firstto describe zonation patterns for zooplankton off Ore-gon. Cross (1964) sampled from 5 to 105 nautical milesfrom shore along 4 transect lines off the ColumbiaRiver, Newport, Coos Bay and the Oregon-Californiaborder. They reported distributions for only a few ofthe most common coastal species and strong zonal gra-dients in abundances were noted for Pseudocalanusminutus (=mimus), Acartia longiremis and Centro-pages mcmurrichi (=abdominalis). For offshore speciesin their samples, they reported only on A. danae andshowed that during summer this species was foundsolely in offshore waters (Cross & Small 1967).

Peterson (1972) conducted 7 cruises on the continen-tal shelf and in oceanic waters off Washington andOregon in summer 1961 and 1962, and was the first toreport distribution patterns of all dominant coastaland oceanic copepod species. Of the common copepodspecies (i.e. species that occurred in >15% of the sam-ples), 2 were found chiefly in shelf waters (Acartia lon-giremis and Pseudocalanus minutus [=mimus]) and 5were found chiefly in oceanic waters (Calanus [=Meso-calanus] tenuicornis, Calocalanus styliremis, Clauso-calanus arcuicornis, C. pergens and Paracalanus spp).

Peterson & Miller (1975) and Peterson et al. (1979)reported on seasonal variations in zonation patterns ofcopepods in continental shelf waters. They described acopepod community during summer which consistedof a nearshore group including Acartia hudsonica andCentropages abdominalis, which was most abundantwithin 5 to 10 km of the shore, and a midshelf groupdominated by A. longiremis, Pseudocalanus spp. andCalanus marshallae, which was found chiefly in mid-shelf and outer shelf waters. Landry & Lorenzen (1989)identified similar species groups for the shelf waters offWashington.

In June or July 1994 through 1998, the Estuarine andOcean Ecology Division of the National Marine FisheriesService conducted plankton and oceanographic surveysoff the coasts of Oregon and Washington in order to de-scribe onshore-offshore variations in the distribution andabundance of northern anchovy Engraulis mordax andPacific sardine Sardinops sagax eggs and larvae, andzooplankton (Bentley et al. 1996). We analyzed a subsetof these zooplankton samples collected in 1994, 1996,and 1997 to examine and describe the onshore-offshorepatterns in copepod community structure off the coast ofOregon. The objectives of this paper include a descrip-tion of copepod community structure (using ordinationtechniques), comparison of on-shelf versus off-shelf com-

munities, and determination of the interannual varia-tions in copepod species dominance using indicator-species analysis. Finally, we relate differences in copepodcommunity structure to oceanographic gradients.

MATERIALS AND METHODS

Field methods. Surveys were conducted with the RV‘Sea Otter’ during 6 to 26 July 1994 and 14 to 25 July1997, and with the RV ‘McArthur’ during 9 to 21 June1996 over a grid of up to 281 stations along up to 14latitudinal transect lines off the coast of Oregon (Fig. 1).The transects extended from 1.9 to 185 km offshore andranged from the Columbia River (46.27° N) to just southof Cape Blanco (42.62° N). Each transect consisted of 17to 20 stations, which began at either 1.9 or 9.3 km fromshore and were evenly spaced at 9.3 km thereafter.

Vertical net tows were conducted using the CALCOFI(California Cooperative Oceanic Fisheries Investiga-tions) protocol (Smith et al. 1985) with the primaryobjective of collecting and quantifying the distributionand abundance of northern anchovy and Pacific sardineeggs and larvae. Tows were made at each station from5 m off the bottom, or from a maximum of 70 m, to thesurface with a 0.25 m diameter CalVET net fitted with150 µm mesh: the filtration area ratio for this net was 9.3,and therefore clogging should not have been a problem(Smith et al. 1985). In addition, we did not see significantamounts of phytoplankton in any of the samples thatwere counted for zooplankton. Plankton samples werepreserved in 5% buffered formalin. Conductivity andtemperature profiles were recorded simultaneously witha Seabird SBE 19 CTD (conductivity, temperature,depth) meter, and salinity and density were later derivedfrom these measurements. Water samples were takenfrom a depth of 3 m with a Niskin bottle, filtered at sea,and analyzed for chl a in the laboratory using standardfluorometric techniques.

For this paper, a total of 186 zooplankton samples wereanalyzed from 4 transects in each year: Columbia River(CR) (46.27° N), Cape Falcon (CF) (45.76° N), Newport(NH) (44.59° N), and Coquille River (CQ) (43.13° N) in1994 and 1996 or the Cape Blanco (CB) (42.62° N) in1997 (Fig. 1). In the laboratory, samples were rinsed infreshwater, poured into graduated cylinders and allowedto settle to determine settled volume. The sample wasthen diluted to 5–10 times the settled volume, suspendedby mixing with a spatula, and aliquots were removedfor counting with a 1.1 ml Hensen-stempel pipette. Twoaliquots were counted, unless more were needed toachieve a total of 300 to 500 animals. Copepods wereidentified to species when possible, or otherwise togenus. Copepods were grouped by copepodite, adultmale or adult female, except for Calanus marshallae, C.

224

Morgan et al.: Onshore-offshore variations in copepod communities

pacificus, Mesocalanus tenuicornis, Euclanus californi-cus, and Neocalanus species, which were also identifiedto copepodite developmental stage. The counts wereconverted into densities (number m–3) using the moutharea of the net and the distance of the tow (m). Biomassof a given species was calculated by multiplying thenumber m–3 by the carbon weight of individuals of agiven developmental stage. Values for carbon are de-rived from literature values or from our own estimates ofweight (see Peterson et al. 2002, Table 2).

Sampling was conducted at stations withoutregard for time of day. Thus, because some cope-pods have been shown to exhibit vertical migra-tion behavior, the possibility of a day and/or nightbias exists. Therefore, we used the Mann-WhitneyU-test to compare the median biomass for dayversus night samples on and off of the shelf. Of the6 comparisons (3 yr × on- or off-shelf), only 1 (1997on-shelf) had a p-value < 0.05 (p = 0.04). For halfof the 6 comparisons, day biomass was higherthan night, and for 3 comparisons night biomasswas higher than day. Therefore, we felt it was rea-sonable to combine the day and night samplestogether for on- versus off-shelf comparisons. Inaddition, several published studies carried out innorthern California Current waters have shownno statistically significant difference in density orbiomass between day and night for the dominantcopepod species which we discuss (Peterson 1972,Shaw & Robinson 1998, Peterson & Keister 2002).

Data reduction and statistical methods. Contourmaps of 1 m temperature and salinity and of cope-pod density and biomass were generated usingkriging as the gridding method in Surfer (GoldenSoftware, V7.0). Distribution and abundance mapsof the individual copepod species were createdusing a layer of shaded raster images of the gridcreated by kriging combined with a classed postmap, which shows the true values at each of theindividual sampling stations where samples werequantified. A linear kriging variogram with ananisotropy ratio of 1 (no anisotropy) and no searchrules was used. All statistical analyses were con-ducted using the computer software PC-ORDV4.36 (McCune & Mefford 1999). For all analyses,copepod densities were first summed for all life-history stages, excluding eggs and nauplii. Alldata were log (x + 1) transformed before statisticalanalysis. The data were then arranged into 4species and 4 environmental matrices (one foreach year and one with all years combined). Thespecies or environmental variables formed thecolumns and stations formed the rows. The size ofthe species matrix was reduced by eliminating‘rare’ taxa (operationally defined as those species

occurring at <5% of the stations for the individualyears and <2% of the stations for all years combined,and which comprised <1.5% of the total population atthe sites where they were present). After transforma-tion, the data were analyzed for outliers. There wereno species in any year with an average distance fromthe mean of more than 2 standard deviations from themean of distances between species.

Nonmetric Multidimensional Scaling (NMDS)(Kruskal 1964, Mather 1976) was used to ordinate

225

Fig. 1. Map of the study area showing location of stations sampled alongthe Oregon coast, USA during 6 to 26 July 1994 (h), 9 to 21 June 1996 ( ), and 14 to 25 July 1997 (D). Temperature and salinity (1 m depth)were recorded at all the oceanographic stations (+). Copepod densitiesand biomass were estimated along 4 hydrographic transects; ColumbiaRiver (CR), Cape Falcon (CF), Newport (NH), and Coquille River (CQ)(1994 and 1996) or Cape Blanco (CB) (1997). The dashed line indicates

the 180 m depth contour (shelf break)

Mar Ecol Prog Ser 249: 223–236, 2003

sample units in species space, to identify station clus-ters with similiar copepod communities, and to relatethem to environmental variables. NMDS has provenuseful in analyzing marine community data (Field et al.1982, Hosie & Cochran 1994, Chiba et al. 2001). NMDSis an iterative search for a ranking and placement of nentities on k dimensions (axes) that minimizes thestress of the k-dimensional configuration. The calcula-tions are based on an n × n distance matrix calculatedfrom the n × p-dimensional main matrix, where n is thenumber of rows and p is the number of columns in themain matrix. ‘Stress’ is a measure of departure frommonotonicity in the relationship between the dissimi-larity (distance) in the original p-dimensional spaceand distance in the reduced k-dimensional ordinationspace. It quantifies between-site dissimilarity andarranges all sites in k-dimensional space, such that therank order of the between-site dissimilarities corre-sponds as closely as possible to the rank order ofbetween-site distances (McCune & Mefford 1999).Thus, the closer that 2 stations are in species space, thecloser the species composition of the stations will be.Runs were made using the Sørensen distance measure(also known as the Bray-Curtis coefficient, Bray & Cur-tis 1957), 4 axes, 150 iterations, 10 runs of real data and30 randomized runs. Plots and output of instability andstress were examined to find the number of axes atwhich the reduction in stress gained by adding anotheraxis was acceptably small (see Mather 1976). In thisdataset, all ordinations were best described by 2 axes(instability < 0.005). The orientation of axes derivedfrom NMDS is arbitrary, and therefore we rotated theplots such that the first axis corresponds to the axis ofmaximum environmental variability.

The environmental data matrices contained values oftemperature (°C), salinity, and density (sigma-t) atdepths of 1 and 30 m, chl a concentration at 3 m depth,longitude and latitude. They also contained groupingvariables for stations located on versus off the conti-nental shelf. The continental shelf group, delineatedby the 180 m isobath, consisted of all stations inshore of25 miles on the CR transect, inshore of 20 miles on theCF and NH transects, and inshore of 10 miles on theCQ and CB transects (Fig. 1).

The nonparametric Multi-Response Permutation Pro-cedure (MRPP) (Mielke 1984, Zimmerman 1985, Mc-Cune & Mefford 1999) was used to test the hypothesis ofno difference in the copepod community composition be-tween samples taken from the on- and off-shelf stations.MRPP is similar to the t-test and the 1-way ANOVAF-test, without requiring the assumptions of normalityand homogeneity of variances. To test for differencesbetween years, MRPP was run separately for on- andoff-shelf stations using year as a grouping variable. TheSørensen distance measure was used in all cases.

Indicator species analysis (ISA) (Dufrene & Legendre1997) was used to define and describe which specieswere typical of on- and off-shelf station groups amongthe 3 yr. The indicator value for a species in a group iscalculated as the product of the relative abundance ofthe species in the group (the mean abundance in thegroup/the sum of the mean abundances in all groups)and the frequency of occurrence of the species in thegroup samples. Indicator values can range from 0 (noindication) to 100 (perfect indication). A score of 100means that the species was present in all samples inthe group and was absent from all samples in othergroups. The highest indicator value for each species istested for statistical significance using a Monte Carlorandomization technique.

RESULTS

Temperature and salinity

During the July 1994 survey, sea surface temperatures(SST) in the sampling area ranged from 8.6 to 16.7°C(Fig. 2). Coastal upwelling was occurring off central andsouthern Oregon as indicated by the presence of water<9°C and salinity greater than 33. During the June 1996survey, SST ranged from 7.2 to 15.4°C. Coastal up-welling was most pronounced off central Oregon, withcold high-salinity water just south of Newport. In theJuly 1997 survey, SST ranged from 9.1 to 18.8°C; thehigher SSTs offshore were the result of a strong down-welling event (Strub & James 2002) likely associatedwith the 1997-1998 El Niño. Coastal upwelling duringthis cruise was weak compared to the other cruises, andagain was most pronounced just south of Newport.

Copepod density and biomass

Table 1 lists the average densities for the dominantspecies in each of the 3 yr. From this table, it is clear thatthe dominant species in on-shelf waters includedCalanus marshallae, Pseudocalanus mimus, Centro-pages abdominalis, and Acartia longiremis. In offshorewaters, the dominant taxa in each of the 3 yr were Meso-calanus tenuicornis, Paracalanus parvus, Clausocalanusspp., Ctenocalanus vanus, and P. mimus. Two otherspecies, Calanus pacificus and Oithona similis, wereabundant in our samples, but were equally abundant onand off the shelf. During 1997, an El Niño year, P. parvuswas the most abundant species in on-shelf waters.

Average copepod biomass during each survey wassimilar among years. For the on-shelf stations, biomass(SE) was 15.9 (2.4), 14.0 (1.4) and 19.6 (2.5) mg C m–3

for 1994, 1996 and 1997 respectively. For the off-shelf

226

Morgan et al.: Onshore-offshore variations in copepod communities

stations, the corresponding biomass values were 5.7(0.6), 7.7 (0.9) and 4.2 (0.4) mg C m–3. The on-shelf andoff-shelf copepod biomasses were significantly differ-ent in all years (Mann-Whitney U-test, p < 0.0001). Bio-mass and density were generally highest in the northand showed strong onshore-offshore gradients (Fig. 3).

Community structure

Using our matrix of species abundance at on- andoff-shelf stations, we tested for differences in communitystructure using MRPP. For this analysis we differentiatedstations by water depth. Stations <180 m deep were de-fined as on-shelf, while stations >180 m deep as off-shelf.We also tested for differences among years. MRPP re-vealed that copepod community structure of on- andoff-shelf stations were significantly different (p < 0.0001)in all years. Copepod community structure was also dif-ferent among years (p < 0.0001) both on and off the shelf.

Ordination of stations in species space

NMDS showed that the stations on and off the conti-nental shelf occupied distinct places in species space,demonstrating (as with MRPP analysis) that there weredistinct differences in copepod community structure(Fig. 4). Monte Carlo analysis gave a probability ofonly 0.03 that these patterns could have been obtainedby chance. The relationship between the distance inthe ordination space and the distance in physical spacewas highly significant (p < 0.0001) (Table 2).

Significant relationships between the ordinationaxis and environmental variables were found. Axis 1was strongly correlated with 30 m temperature anddensity, longitude, and 1 m temperature (Table 2).Higher temperatures were associated with off-shelfsamples, while denser waters were associated withthe on-shelf samples (Fig. 4). Chlorophyll a was alsohighly correlated with Axis 1 in 1994 and 1997 but notin 1996.

227

Fig. 2. Sea surface (1 m depth) temperatures (°C) and salinities along the Oregon coast, USA during 6 to 26 July 1994, 9 to 21 June 1996 and 14 to 25 July 1997. The dashed line indicates the 180 m depth contour

Mar Ecol Prog Ser 249: 223–236, 2003228

Table 1. Average densities (no. m–3) and SE (in parentheses) of copepod species in waters on and off the continental shelf off Ore-gon in summer 1994, 1996, and 1997. The number of samples in each category is in parentheses beneath the respective heading.

The 4 species in bold are the dominant on-shelf species

Species 1994 1996 1997On-shelf Off-shelf On-shelf Off-shelf On-shelf Off-shelf

(14) (45) (18) (38) (17) (54)

Calanus marshallae Frost 267 (108) 38 (15) 169 (28) 21 (12) 110 (21) 1 (4)Calanus pacificus Brodsky 19 (11) 42 (11) 14 (6) 17 (4) 26 (10) 17 (22)Mesocalanus tenuicornis (Dana) 46 (12) 10 (3) 3 (2) 15 (20)Eucalanus californicus Johnson 1 (1) 1 (1) 2 (1) 14 (4) 9 (13)Paracalanus parvus (Claus) 126 (33) 467 (69) 73 (20) 501 (91) 3315 (874) 863 (872)Calocalanus styliremis Giesbrecht 12 (3) 2 (1) 30 (6) 0.3 (0.3) 49 (52)Clausocalanus spp. 7 (5) 97 (19) 11 (5) 207 (56) 39 (19) 87 (92)Ctenocalanus vanus Giesbrecht 63 (13) 3 (2) 131 (25) 11 (8) 133 (109)Pseudocalanus mimus (Frost) 1457 (240) 435 (80) 1295 (125) 707 (201) 1346 (196) 161 (269)Metridia pacifica Brodsky 9 (5) 22 (5) 72 (35) 51 (10) 22 (8) 14 (28)Centropages abdominalis Sato 61 (22) 7 (2) 15 (7) 12 (9) 540 (218) 13 (36)Candacia bipinnata (Giesbrecht) 1 (1) 2 (1) 6 (9)Acartia spp. unidentified. copepodites 473 (113) 132 *(36) 67 (23) 22 (16) 147 (42) 18 (34)Acartia danae Giesbrecht 4 (1) 0.2 (0.2) 8 (13)Acartia longiremis (Lilljeborg) 974 (214) 112 (54) 80 (24) 17 (10) 199 (48) 5 (14)Acartia tonsa Dana 5 (2) 3 (1) 1 (1) 0.3 (0.4) 4 (3) 1 (2)Acartia hudsonica Pinhey 9 (5) 1 (1) 7 (5) 17 (7)Microsetella sp. 2 (1) 2 (1) 1 (1) 21 (4) 1 (1) 13 (12)Corycaeus anglicus Lubbock 2 (2) 9 (2) 11 (5) 26 (11) 24 (7) 10 (16)Oithona similis Claus 1677 (273) 1865 (131) 1305 (180) 1872 (233) 1730 (191) 1549 (824)All other copepod species 2 (2) 12 (2) 6 (2) 18 (2) 15 (3) 9 (1)Total copepods 618 (65) 303 (25) 364 (39) 299 (23) 721 (109) 233 (18)

Fig. 3. Copepod density (no. m–3) and biomass (mg m–3) estimates off the Oregon coast, USA from stations sampled (+) along 4 hy-drographic transects: Columbia River (CR), Cape Falcon (CF), Newport (NH), and Coquille River (CQ) during 6 to 26 July 1994 and9 to 21 June 1996; CR, CF, NH, and Cape Blanco (CB) during 14 to 25 July 1997. The dashed line indicates the 180 m depth contour

Morgan et al.: Onshore-offshore variations in copepod communities

The NMDS ordination of stations classified by waterdepth revealed good separation between on-shelf sta-tions (solid symbols; clustered on the left of Fig. 4) andoff-shelf stations (open symbols; clustered on the rightside of Fig. 4). There was, however, some overlap forstations in all of the depth bins. Though 180 m waterdepth did not always define the break point in commu-nity composition, data from stations less than 300 mdeep did divide the 2 communities in almost everycase. This suggests that the true break in the copepodcommunity composition is closely associated with thelocation of the continental shelf break, but is not astep-like change.

Indicator species analysis

ISA revealed similarities amongyears for which copepod species wereindicators of shelf waters (Table 3).Three species were consistent indica-tors of on-shelf copepod communitiesin all years: Calanus marshallae, Pseu-docalanus mimus and Acartia longi-remis. Two additional species, Centro-

pages abdominalis and A. hudsonica, were indicatorsof on-shelf waters in 1994 and 1997. In 1997, Para-calanus parvus was also a good indicator of on-shelfwaters.

For off-shelf waters, the species which were goodindicators differed among years. As many as 12 spe-cies were identified as good indicators of off-shelf con-ditions in one year or another, but only 4 speciesappeared as consistent indicators each year; Meso-calanus tenuicornis, Calocalanus styliremis, Clauso-calanus spp. and Ctenocalanus vanus (Table 3). Twoadditional species were off-shelf indicators in 1994 and

229

Fig. 4. Results of nonmetric multidimensional scaling (NMDS) of1994, 1996 and 1997 stations in species space. Stations arecoded by depth range (m), and all environmental variables witha significant correlation (r2 > 0.35) with Axis 1 are shown. Thedirection of the arrow indicates the direction of the correlationwith the stations and the fraction (%) of variance in the copepoddata explained by that parameter (see Table 2). All ordinationsare rotated towards 30 m temperature, the environmental vari-able that most often explained the largest fraction of the vari-ance. The dashed line indicates the position along the horizon-tal axis beyond which no more on-shelf stations occur. Off-shelfstations that fell to the left of this line are labeled. Station labels

indicate transect and nautical miles offshore

Table 2. The coefficients of determination (r2) for the correlation between ordi-nation distances and original distances and for the maximum correlation of 30 mtemperature, 30 m denisty, longitude, 1 m temperature, and chl a with Axis 1

Year Coefficients of determination (r2)(n) Ordination 30 m 30 m Longi- 1 m Chl a

vs temp- density tude temp- (n)original erature erature

1994 (59) 0.86 0.87 0.73 0.62 0.53 0.36 (54)1996 (56) 0.91 0.77 0.46 0.81 0.52 0.02 (24)1997 (71) 0.90 0.77 0.67 0.68 0.26 0.45 (37)All yearscombined (186) 0.84 0.76 0.56 0.63 0.39 0.18 (115)

Mar Ecol Prog Ser 249: 223–236, 2003

1996; Paracalanus parvus and Oncaea spp., 2 otherswere off-shelf indicators in 1996 and 1997; Eucalanuscalifornicus and Microsetella spp, and 4 other specieswere off-shelf indicators in only one of the cruises (seeTable 3). In every case, the off-shelf indicator speciesin question are known to be either subtropical neriticor Transition Zone species (Fleminger 1976). Two ofthese species, M. tenuicornis and C. vanus, are 2 of themost frequently occurring species in the central andsouthern California Current (Fleminger 1967).

Distributions of dominant species

Spatial variations in the distribution of the 3 domi-nant subarctic coastal species (Calanus marshallae,Pseudocalanus mimus, and Acartia longiremis) and 3offshore subtropical and/or Transition Zone species(Mesocalanus tenuicornis, Paracalanus parvus, andCtenocalanus vanus) are shown in Figs. 5 to 7. Wechose to illustrate these species because they areabundant in all samples (Table 1), and ISA (Table 3)revealed them to be excellent indicators of coastalwaters and oceanic waters, respectively. P. mimus hadhigh densities at most on-shelf stations (Fig. 5). High-est densities of C. marshallae (Fig. 5) and A. longiremis(Fig. 6) were seen at the on-shelf stations centeredoff Newport in 1994. In 1996 and 1997, A. longiremiswas most abundant on the Columbia River and CapeFalcon transects, whereas C. marshallae was evenlydistributed along the coast. The off-shelf species wereusually most abundant along the northern or centraltransect lines (Figs. 6 & 7). However, during 1997, P.parvus (Fig. 6) had its highest densities in on-shelfwaters off the Columbia River and Cape Falcon. M.tenuicornis (Fig. 7) was most abundant offshore and tothe north in all years. C. vanus (Fig. 7) was most abun-

dant offshore and to the north in 1997, but was moreevenly distributed in the other 2 yr.

The NMDS ordination shown in Fig. 4 was overlaidwith the densities of 3 closely related Calanidae spe-cies to explore the on-/off-shelf differences (Fig. 8).This analysis showed that Calanus marshallae clus-tered as a coastal species, Mesocalanus tenuicornis asan off-shelf species, and C. pacificus was distributedthroughout both regions (Fig. 8). We compared the 4Acartia species in the same manner (not illustrated),and found that A. hudsonica was present chiefly at theshallowest stations on the shelf, A. longiremis in mid-and outer-shelf samples, and A. danae only at thedeepest water stations. A. tonsa occurred in only 19samples and was equally common at both on- andoff-shelf stations.

DISCUSSION

One purpose of this work was to test the hypothesisof on-/off-shelf differences in the biomass, speciescomposition and community structure of copepods inthe upwelling zone off Oregon. We found that copepodbiomass was similar among years and averaged16.5 mg C m–3 (on-shelf, range 14.0 to 19.6 mg C m–3)and 5.9 mg C m–3 (off-shelf, range 4.2 to 7.7 mg C m–3),with on-shelf biomass greater than off-shelf biomassby a factor of 2.8. The chief reason for this difference inbiomass may be related to differences in primary pro-duction rates and chlorophyll concentration. The aver-age chl a concentrations during our cruises were 4.4 µgl–1 on-shelf and 0.6 µg l–1 off-shelf, different by a factorof 7. Perry et al. (1989) also reported higher primaryproduction rates on-shelf compared to off-shelf, aver-aging 0.95 and 0.33 mg C m–2 d–1 respectively forstations off the Washington coast, different by a factor

230

Table 3. Copepod indicator species (p < 0.05) for summer 1994, 1996, and 1997 off the Oregon coast. Species are listed by decreasing order of indicator value within each region

Region 1994 indicator species 1996 indicator species 1997 indicator species

On-shelf Acartia longiremis Calanus marshallae Calanus marshallaeCentropages abdominalis Acartia longiremis Acartia longiremisPseudocalanus mimus Pseudocalanus mimus Centropages abdominalisCalanus marshallae Pseudocalanus mimusAcartia hudsonica Paracalanus parvus

Acartia hudsonica

Off-shelf Clausocalanus spp. Oncaea spp. Calocalanus styliremisCtenocalanus vanus Clausocalanus spp. Ctenocalanus vanusParacalanus parvus Ctenocalanus vanus Microsetella sp.Mesocalanus tenuicornis Calocalanus styliremis Clausocalanus spp.Calocalanus styliremis Microsetella sp. Eucalanus bungii californicusOncea sp. Mesocalanus tenuicornis Mesocalanus tenuicornisMetridia pacifica Paracalanus parvus Acartia danaeCorycaeus anglicus Eucalanus bungii californicus Candacia bipinnata

Morgan et al.: Onshore-offshore variations in copepod communities

of 3. Thus, higher copepod biomass and copepod pro-duction in on-shelf waters would be expected. Thiswas confirmed during the June 1996 cruise, duringwhich Peterson et al. (2002) measured copepod growthrates at 37 on-shelf stations and 36 off-shelf stationsand calculated copepod production. They found thatcopepod production averaged 2.3 mg C m–3 d–1 for theon-shelf stations and 0.8 mg C m–3 d–1 for the off-shelfwaters, differing by a factor of 2.9. Thus, differencesbetween on- and off-shelf plankton biomass andproduction were similar; phytoplankton biomass (aschlorophyll), primary production, copepod biomassand secondary production were all greater in on-shelfwaters than off-shelf waters by a factor of ≥3.

Not only were there on-/off-shelf variations in cope-pod biomass, but we also observed differences incopepod community structure. We suggest that thesedifferences in both copepod biomass and species com-position derive from the different origins of the source

waters for on- and off-shelf habitats. For example, 3copepod species were consistently more abundant onlyat the on-shelf stations: Calanus marshallae, Pseudo-calanus mimus, and Acartia longiremis. We know thatthese species dominate shelf waters off Newport, Ore-gon (Peterson & Miller 1975) as well as coastal watersto the north; Vancouver Island (Mackas et al. 2001),Gulf of Alaska (Coyle et al. 1990), and the Bering Sea(Johnson & Brinton 1963). Thus, these 3 species aregood indicators of the presence of Gulf of Alaskacoastal water on the Oregon shelf.

On the other hand, offshore California Currentwaters are dominated by a mixture of copepod species,some from the Transition Zone (Tables 1 and 3), andothers that are typical of the temperate-subtropicalcoastal neritic regions of central and southern Cali-fornia. The Transition Zone is a band of water thatextends between about 40 and 45° N across the PacificOcean from Japan to Oregon and northern California.

231

Fig. 5. Density estimates of Pseudocalanus mimus and Calanus marshallae off the Oregon coast, USA from stations (+) along 4 hy-drographic transects: Columbia River (CR), Cape Falcon (CF), Newport (NH), and Coquille River (CQ) during 6 to 26 July 1994 and9 to 21 June 1996; CR, CF, NH, and Cape Blanco (CB) during 14 to 25 July 1997. The dashed line indicates the 180 m depth contour

Mar Ecol Prog Ser 249: 223–236, 2003

Copepod species that are typical of the Transition Zoneinclude Mesocalanus tenuicornis, Calanus pacificus,Eucalanus californicus and Candacia bipinnata (Fle-minger 1967). Subtropical neritic species are also com-monly found in offshore waters during summer andinclude Paracalanus parvus, Ctenocalanus vanus,Acartia tonsa and Corycaeus anglicus (Fleminger1967, Barnett & Jahn 1987). These species are moretypically found in coastal waters of central and south-ern California, but they prosper in deep waters offOregon. We commonly found Clausocalanus specieswith the warm-water group. The systematics of thisgenus are difficult, and we did not attempt to routinelyidentify them to species. However, detailed analysis ofadditional samples collected at biweekly intervals atseveral on-shelf stations off Newport, Oregon from1996 to present, reveal that the dominant species areC. arcuicornis, C. pergens and C. paragergens (Peter-

son et al. 2002). These species are also known to beTransition Zone species (Frost & Fleminger 1968).Thus, though the offshore copepod community hasmixed affinities, all species can be classified as ‘warmwater’ species. These observations collectively supportthe physical oceanographic paradigm of Hickey (1998)that waters offshore of the continental shelf in thenorthern California Current region have their originsin the Transition Zone and the west wind drift.

It is noteworthy that the ‘offshore community’ con-sisting of subtropical and Transition Zone copepodspecies is virtually the same copepod community thatoccurs in coastal waters off Oregon and Washingtonduring winter months (Peterson & Miller 1977, theirTable 1). During winter months, southwesterly stormstransport offshore waters (and offshore copepods) on-shore, resulting in downwelling at the coast. In addi-tion, the northward-flowing coastal Davidson Current

232

Fig. 6. Density estimates of Acartia longiremis and Paracalanus parvus off the Oregon coast, USA from stations (+) along 4 hydro-graphic transects: Columbia River (CR), Cape Falcon (CF), Newport (NH), and Coquille River (CQ) during 6 to 26 July 1994 and 9to 21 June 1996; CR, CF, NH, and Cape Blanco (CB) during 14 to 25 July 1997. The dashed line indicates the 180 m depth contour

Morgan et al.: Onshore-offshore variations in copepod communities

advects warm-water subtropical neritic copepods (suchas Paracalanus parvus, Ctenocalanus vanus andCorycaeus anglicus) from the California shelf toOregon shelf waters. We speculate that these warmwater neritic copepod species are found offshore ofOregon during summer because they are advected off-shore in spring when upwelling commences, and sub-sequently remain in the offshore area during springand summer months because southward flows in theoffshore California Current are sluggish (Hickey 1998).This explains why the offshore copepod community offOregon is a mixture of species of different origins,some are transported from the west wind drift, where-as others are permanent residents of the northernCalifornia Current that live offshore in summer but incontinental shelf waters in winter.

We found that the geographic location between theon- and off-shelf community occurred at or just off-

shore of the shelf break. The fidelity of the separationof on- and off-shelf communities was very high, withmore than 94% of our stations grouped along anonshore-offshore environmental gradient. The few sta-tions that did not group exactly with other members oftheir community are indicated on Fig. 3 by their station(transect letter and distance in nautical miles offshore).Misclassification occurred where upwelling was par-ticularly strong in association with a headland: CapeFalcon in 1996 and 1997, and Cape Blanco (StationsCQ 15 in 1994, and CB 10 and 20 in 1997. Capes areregions of high offshore transport, particularly CapeBlanco (Barth et al. 2000). It is common for coastalcopepods to be transported offshore in that region(Peterson & Keister 2002). Another example of offshoretransport was in 1996 along the Columbia River tran-sect, where offshore transport of coastal species by theColumbia River plume probably occurred.

233

Fig. 7. Density estimates of Mesocalanus tenuicornis and Ctenocalanus vanus off the Oregon coast, USA from stations (+) along 4 hy-drographic transects: Columbia River (CR), Cape Falcon (CF), Newport (NH), and Coquille River (CQ) during 6 to 26 July 1994 and9 to 21 June 1996; CR, CF, NH, and Cape Blan (CB) during 14 to 25 July 1997. The dashed line indicates the 180 m depth contour

Mar Ecol Prog Ser 249: 223–236, 2003

Hickey (1998) emphasized that variability in currentsover the continental shelf off Washington and Oregonhave different time and space scales from currentsseaward of the shelf. Moreover, the degree to whichthe off-shelf California Current interacts with coastalwaters is unclear (Hickey 1989, p. 43–45). There areevidently no sharp boundaries in hydrographic proper-ties between waters, and rather a gradual transitionfrom cold salty water on the shelf to warm fresherwater off the shelf. Recent drifter observations suggestthat, at least in the upper layers of the water column,the interactions between the shelf and slope watersare quite weak (J. Barth unpubl. data, available athttp://diana.coas.oregonstate.edu/drift/recent.shtml).Drifters launched in on-shelf waters off Newport dur-ing summer invariably move south at high speeds andremain in on-shelf waters until at least the latitude ofCape Blanco (42.5° N). However, drifters launched inslope waters (40 to 50 km from shore) often move to thewest or meander about the area. Thus, we suggest that

the apparent lack of interaction between the 2 watertypes may partially explain why we observed 2 distinctcopepod communities.

Another related mechanism that could explainseparation of the 2 copepod communities is related tooffshore transport during coastal upwelling. Activecoastal upwelling does not extend beyond the shelfbreak off central Oregon (Huyer 1983) thus coastalspecies will not normally be transported beyond theshelf. Earlier papers (Peterson et al. 1979, Peterson1998) found that coastal copepod species are retainedin coastal upwelling zones through ontogenetic migra-tion. We suggest that the retention mechanisms thatkeep coastal species on the shelf may explain thestrong contrasts in copepod community structure be-tween on-shelf and off-shelf waters. That is, in theregion of central Oregon, coastal copepod species arenot transported offshore beyond the shelf break duringupwelling, and upwelling prevents offshore speciesfrom being transported onshore.

234

Fig. 8. Results of non-metric multidimensional scaling (NMDS) of 1994, 1996 and 1997 stations in species space. Stations arecoded by whether they occur on-shelf (≤180 m depth) or off-shelf (>180 m depth). The size of the symbol is relative to the densityof Calanus marshalle, C. pacificus, or Mesocalanus tenuicornis at that station, with the smallest symbol representing none of thatspecies present. All ordinations were rotated towards longitude. Axis scales in NMDS are relative, based on non-metric ranking

of similarity and therefore not shown

Morgan et al.: Onshore-offshore variations in copepod communities

Another possible mechanism accounting for theshelf-oceanic gradient in copepod community structureis related to the relative ability of species to survive in adifferent water types. That is, should there be an ex-change of shelf and slope waters through some physicalprocess, oceanic species may not survive in coastalwaters and vice versa. Oceanic species are adapted torelatively particle-free oligotrophic water, and may beincapable of feeding in particle-rich waters with chloro-phyll values in excess of 10 to 20 µg chl a l–1, typical forsummertime off the Oregon and Washington coasts.The shelf environment is dominated by diatoms whilethe offshore oligotrophic environment is dominated bynannophytoplankton (Evelyn Sherr pers. comm.). Dif-ferences in availability of food type and size may ex-plain why coastal species prosper in coastal regions andsubtropical species prosper in oligotrophic waters, butnot vice versa. There are strong temperature gradientsbetween shelf and slope waters. The cool waters on theshelf might prevent successful colonization of shelf wa-ters by offshore, warm-adapted, species. Similarly,cold-adapted shelf species may also be unable to sur-vive in relatively warm and particle-free oligotrophicwaters offshore of the shelf. Finally, predator guildsare different between coastal and oceanic waters, withcoastal waters dominated by ctenophores, scypho-medusae, and zooplanktivorous fishes (osmeriids andclupeids) and oceanic waters by Pacific saury Cololabissaira, mesopelagic fishes (at night), and euphausiids.Thus, a variety of biological interactions coupled withweak offshore Ekman transport and weak interaction ofshelf and slope waters may also explain the observedzonation in copepod communities.

Our observation of a separation between an on- andoff-shelf copepod community in association with the con-tinental shelf break appears to be a general result, beingcommon to both upwelling and non-upwelling regions.For example, separation of water types and copepodcommunity types occurs at or near the shelf break inmost coastal upwelling regions: Baja California (Long-hurst 1967, Cervantes-Duarte et al. 1993), Peru (San-tander 1981), southwest Brazil off Cabo Frio (Lopes etal. 1999), and Cap Blanc, Mauritania (Kuipers et al.1993). To cite only a few examples from non-upwellingregions, Fernández-Alamo et al. (2000) found the limitbetween a neritic and oceanic copepod assemblagecorresponded largely with the border of the continentalshelf in the Gulf of Tehuantepec, as did Seridji &Hafferssas (2000) for the Mediterranean Sea off the coastof Algeria, Siokou-Frangou et al. (1998) in the AegeanSea, and Grice & Hart (1962) for the continental shelf offNew York. On the other hand, shelf-slope zonation incopepod community structure is not observed in up-welling systems with strong wind forcing and a verynarrow shelf such as off northern Chile (23° N) (Escrib-

ano & Hildalgo 2000), and Somalia (NE Africa) (Smith1982). In these systems, coastal animals are displacedgreat distances from shore into off-shelf waters.

In conclusion, the copepod communities off the Ore-gon coast closely resemble the structure of the domi-nant water masses. Species in coastal waters reflecttheir Gulf of Alaska origins, whereas species in off-shore waters are warm water species of mixed origins.Clearly, some species are advected into the offshoreregion from the west wind drift and Transition Zone.Others, are warm water neritic species with origins inshelf waters of central and southern California. Thesewarm water neritic species are advected to Oregon’scoastal waters in winter by the Davidson Current.We speculate that during spring and summer, thesewarm-water species are advected to offshore waters bycoastal upwelling, where they establish viable popula-tions because equatorward advection is slow. Wesuggest that the coastal copepod community remainsseparated from the offshore community through acombination of factors, including weak physical inter-action between shelf and offshore waters, and lack ofdirect exchange of coastal waters with oceanic waters.

Acknowledgements. We thank the captain and crew of theRVs ‘Sea Otter’ and ‘McArthur’ for their cooperation andassistance during these cruises. We would especially like tothank P. J. Bentley for his participation and hard work duringall of the cruises. This project was partially funded by aNOAA/Recovery Protected Species grant to R.L.E. We thankJ. Keister, Dr. C. B. Miller, and 3 referees for comments on themanuscript.

LITERATURE CITED

Barnett AM, Jahn AE (1987) Pattern and persistence of anearshore planktonic ecosystem off Southern California.Cont Shelf Res 7(1):1–25

Barth JA, Pierce SD, Smith RL (2000) A separating coastalupwelling jet at Cape Blanco, Oregon and its connectionto the California current system. Deep-Sea Res II 47(5-6):783–810

Bentley PJ, Emmett RL, Lo NCH, Moser G (1996) Egg pro-duction of Pacific sardine (Sardinops sagax) off Oregon in1994. CalCOFI Rep 37:193–200

Bray JR, Curtis JT (1957) An ordination of the upland forestcommunities in southern Wisconsin. Ecol Monogr 27(4):325–349

Cervantes-Duarte R, Aguniga Garcia S, Hernandez Trujillo S(1993) Upwelling conditions associated to the distributionof zooplankton in San Hipolito, BCS. Ciencias Mar 19(1):117–135

Chiba S, Ishimaru T, Hosie GW, Fukuchi M (2001) Spatio-temporal variability of zooplankton community structureoff east Antarctica (90 to 160° E). Mar Ecol Prog Ser 216:95–108

Coyle KO, Paul AJ, Ziemann DA (1990) Copepod populationsduring the spring bloom in an Alaskan subarctic embay-ment. J Plankton Res 12(4):759–797

Cross FA (1964) Seasonal and geographical distribution of

235

Mar Ecol Prog Ser 249: 223–236, 2003

pelagic copepods in Oregon coastal waters. MS thesis,Oregon State University, Corvallis

Cross FA, Small LF (1967) Copepod indicators of surfacewater movements off the Oregon coast. Limnol Oceanogr12:60–72

Dufrene M, Legendre P (1997) Species assemblages andindicator species: the need for a flexible asymmetricalapproach. Ecol Monogr 67(3):345–366

Escribano R, Hidalgo P (2000) Spatial distribution of copepodsin the north of the Humbolt Current region off Chile dur-ing coastal upwelling. J Mar Biol Assoc UK 80(2):283–290

Fernández-Alamo MA, Sanvicente-Anorve L, Alameda-De-La-Mora G (2000) Copepod assemblages in the Gulf ofTehuantepec, Mexico. Crustaceana 73(9):1139–1153

Field JG, Clarke KR, Warwick RM (1982) A practical strategyfor analyzing multispecies distribution patterns. Mar EcolProg Ser 8:37–52

Fleminger A (1967) Distributional atlas of calanoid copepodsin the California Current region, Part II. California Coop-erative Oceanic Fisheries Investigation Atlas No. 7, xvi +1–213 plates

Frost BW, Fleminger A (1968) A revision of the genus Clauso-calanus (Copepoda: Calanoida) with remarks on distribu-tional patterns in diagnostic characters. Bull Scripps InstOceanogr 12:1–235

Grice GD, Hart AD (1962) The abundance, seasonal occur-rence and distribution of the epizooplankton betweenNew York and Bermuda. Ecol Monogr 32:287–309

Hickey BM (1989) Pattern and process of circulation over theWashington continental shelf and slope. In: Landry MR,Hickey BM (eds) Coastal oceanography of Washingtonand Oregon. Elsevier Oceanogr Ser 47:41–115

Hickey BM (1998) Coastal Oceanography of western NorthAmerica from the tip of Baja California to Vancouver Island(coastal segment 8E). In: Robinson AR, Brink KH (eds) Thesea: ideas and observations on progress in the study of theseas. The global coastal ocean: regional studies and syn-theses. Interscience Publishers, New York, p 345–393

Hosie GW, Cochran RG (1994) Mesoscale distribution pat-terns of macrozooplankton communities in Prydz Bay,Antarctica-January to February 1991. Mar Ecol Prog Ser106:21–39

Huyer A (1983) Coastal upwelling in the California Currentsystem. Prog Oceanogr 12:259–284

Johnson MW, Brinton E (1963) Biological species, water-masses and currents. In: Hill M (ed) The sea, Vol 2. Inter-science Publishers, New York, p 381–414

Kruskal JB (1964) Nonmetric multidimensional scaling: anumerical method. Psychometrika 29:1–27

Kuipers BR, Witte HJ, Gonzalez SR (1993) Zooplankton distri-bution in the coastal upwelling system along the Bancd’Arguin, Mauritania. Hydrobiologia 258:133–149

Landry MR, Lorenzen CJ (1989) Abundance, distribution, andgrazing impacts of zooplankton on the Washington shelf.In: Landry MR, Hickey BM (eds) Coastal Oceanography ofWashington and Oregon. Elsevier Oceanogr Ser 47:175–210

Longhurst A (1967) Diversity and trophic structure of zoo-plankton communities in the California Current. Deep-Sea Res 14:393–408

Lopes RM, Brandini FP, Gaeta S (1999) Distribution patternsof epipelagic copepods off Rio de Janeiro (SE Brazil) insummer 1991/1992 and winter 1992. Hydrobiologia 411:161–174

Mackas DL, Thomson RC, Galbraith M (2001) Changes in thezooplankton community of the British Columbia continen-tal margin, 1985–1999, and their covariation with oceano-graphic conditions. Can J Fish Aquat Sci 58:685–702

Mather PM (1976) Computational methods of multivariateanalysis in physical geography. J Wiley & Sons, London

McCune B, Mefford MJ (1999) Multivariate analysis of eco-logical data, Version 4.36 beta. MjM Software, GlenedenBeach, OR

Mielke PW Jr (1984) Meterological applications of permuta-tion techniques based on distance functions. In: Krish-naiah PR, Sen PK (eds) Handbook of statistics, Vol 4.Nonparametric methods. Elsevier Science Publishers,Amsterdam, p 813–830

Perry MJ, Bolger JP, English DC (1989) Primary productionin Washington coastal waters. Elsevier Oceanogr Ser 47:117–138

Peterson W (1998) Life cycle strategies of copepods in coastalupwelling zones. J Mar Syst 15(1-4):313–326

Peterson WK (1972) Distribution of pelagic Copepoda off thecoasts of Washington and Oregon during 1961 and 1962.In: Pruter AT, Alverson DL (eds) The Columbia River Estu-ary and adjacent ocean waters. University of WashingtonPress, Seattle, p 313–343

Peterson WT, Keister JE (2002) The effect of a large cape ondistribution patterns of coastal and oceanic copepods offOregon and northern California during the 1998-1999 ElNiño-La Niña. Prog Oceanogr 53(2-4):389–411

Peterson WT, Miller CB (1975) Year-to-year variations in theplanktology of the Oregon upwelling zone. Fish Bull 73(3):642–653

Peterson WT, Miller CB (1977) Seasonal cycle of zooplanktonabundance and species composition along the centralOregon coast. Fish Bull 75(4):717–724

Peterson WT, Miller CB, Hutchinson A (1979) Zonation andmaintenance of copepod populations in the Oregon up-welling zone. Deep-Sea Res 26A:467–494

Peterson WT, Gómez-Gutiérrez J, Morgan CA (2002) Cross-shelf variation in calanoid copepod production during sum-mer 1996 off the Oregon coast, USA. Mar Biol 141:353–365

Peterson WT, Keister JE, Feinberg L (2002) The effects of the1997–1999 El Niño/La Niña events on hydrography andzooplankton off the central Oregon coast. Prog Oceanogr54(1-4):381–398

Santander BH (1981) The zooplankton in an upwelling areaoff Peru. In: Richards FA (ed) Coastal upwelling, coastaland estuarine sciences (1). American Geophysical Union,Washington, DC, p 411–416

Seridji R, Hafferssas A (2000) Copepod diversity and communitystructure in the Algerian Basin. Crustaceana 73(1):1–23

Shaw W, Robinson CLK (1998) Night versus day abundanceestimates of zooplankton at two coastal stations in BritishColumbia, Canada. Mar Ecol Prog Ser 175:143–153

Siokou-Frangou I, Papahanassiou E, Lepretre A, Frontier S(1998) Zooplankton assemblages and influence of envi-ronmental parameters on them in a Mediterranean coastalarea. J Plankton Res 20(5):847–870

Smith PE, Flerx W, Hewitt RP (1985) The CalCOFI verticalegg two (CalVET) net. In: Lasker (ed) An egg productionmethod for estimation spawning biomass of pelagic fish:application to the northern anchovy (Engraulis mordax).US Dept Comm, NOAA Tech Rep NMFS 36:27–32

Smith SL (1982) The northwestern Indian Ocean during themonsoons of 1979: distribution, abundance, and feeding ofzooplankton. Deep-Sea Res 29(11A):1331–1353

Strub PT, James C (2002) The 1997–1998 oceanic El Niño sig-nal along the southeast and northeast Pacific boundaries-an altimetric view. Prog Oceanogr 54(1-4):439–458

Zimmerman GM (1985) Use of an improved statistical methodfor group comparisons to study effects of prairie fire.Ecology 66(2):606–611

236

Editorial responsibility: Michael Landry (ContributingEditor), Honolulu, Hawaii, USA

Submitted: August 30, 2001; Accepted: October 30, 2002 Proofs received from author(s): February 24, 2003